Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Gene 265 (2001) 61±69 www.elsevier.com/locate/gene LHX3 transcription factor mutations associated with combined pituitary hormone de®ciency impair the activation of pituitary target genes Kyle W. Sloop 1, Gretchen E. Parker 1, Kimberly R. Hanna, Heather A. Wright, Simon J. Rhodes* Department of Biology, Indiana University-Purdue University Indianapolis, 723 West Michigan Street, Indianapolis, IN 46202-5132, USA Received 15 December 2000; received in revised form 9 January 2001; accepted 24 January 2001 Received by A.J. van Wijnen Abstract The Lhx3 LIM homeodomain transcription factor is critical for pituitary gland formation and speci®cation of the anterior pituitary hormone-secreting cell types. Two mutations in LHX3, a missense mutation changing a tyrosine to a cysteine and an intragenic deletion that results in a truncated protein lacking the DNA-binding homeodomain, have been identi®ed in humans. These mutations were identi®ed in patients with retarded growth and combined pituitary hormone de®ciency and also abnormal neck and cervical spine development. For both the LHX3a and LHX3b isoforms, we compared the ability of wild type and mutant LHX3 proteins to trans-activate pituitary genes, bind DNA recognition elements, and interact with partner proteins. The tyrosine missense mutation inhibits the ability of LHX3 to induce transcription from selected target genes but does not prevent DNA binding and interaction with partner proteins such as NLI and Pit-1. Mutant LHX3 proteins lacking a homeodomain do not bind DNA and do not induce transcription from pituitary genes. These studies demonstrate that mutations in the LHX3 isoforms impair their gene regulatory functions and support the hypothesis that defects in the LHX3 gene cause complex pituitary disease in humans. q 2001 Elsevier Science B.V. All rights reserved. Keywords: Alpha glycoprotein subunit; Growth; Homeodomain; LIM; Prolactin 1. Introduction The anterior pituitary gland is the central endocrine regulator of growth, metabolism, lactation, reproduction, and the response to stress. These processes are controlled by the actions of hormones released from the distinct cell types that comprise this organ. These cell types are somatotropes that secrete GH, thyrotropes that release TSH, lactotropes that produce PRL, gonadotropes that synthesize LH and FSH, and corticotropes that secrete ACTH. Following early inductive signals, these cells differentiate from a common origin in a spatial and temporal manner that is controlled by the actions of pituitary-speci®c and pituitary-enriched transcription factors (reviewed in Burrows et Abbreviations: ACTH, adrenocorticotropin; aGSU, alpha glycoprotein subunit; CPHD, combined pituitary hormone de®ciency; EMSA, electrophoretic mobility shift analysis; FSH, follicle-stimulating hormone; GH, growth hormone; GST, glutathione-S-transferase; LH, luteinizing hormone; LHX3, LIM homeobox gene 3; LIM-HD, LIM homeodomain; NLI, nuclear LIM interactor; PCR, polymerase chain reaction; PRL, prolactin; TSH, thyroid-stimulating hormone * Corresponding author. Tel.: 11-317-278-1797; fax: 11-317-274-2846. E-mail address: [email protected] (S.J. Rhodes). 1 These authors contributed equally to this study. al., 2000). These regulatory factors include the Lhx3, Lhx4, Hesx1, Pitx-1, Pitx-2, Prop-1, and Pit-1 homeodomain proteins (Burrows et al., 2000). Mutations in the genes encoding several of these factors have been demonstrated to cause pituitary disease. Recessive and dominant mutations of the PIT1 gene lead to the development of a hypoplastic anterior pituitary gland and CPHD (reviewed in Parks and Brown, 1999). These patients display loss of circulating GH and PRL and often have low or absent levels of TSH. Several types of mutations in PIT-1 have been shown to impair its function. For example, the A158P (PfaÈf¯e et al., 1992) and P239S (Pernasetti et al., 1998) mutations reduce the ability of PIT-1 to induce transcription of target genes. Nonsense mutations such as R172X (Tatsumi et al., 1992) and E250X (Irie et al., 1995) result in shortened proteins that do not contain the DNA-binding homeodomain. The R271W defect may exert a dominant negative effect by enabling this mutant of PIT-1 to bind certain DNA elements better than wild type PIT-1 (e.g. Cohen et al., 1995). This mutation also may prevent dimerization of the molecule (Jacobson et al., 1997). To date, all mutations described in the PROP-1 gene are recessive (reviewed in Parks and Brown, 1999). PROP-1 defects cause a similar form of CPHD to that observed in PIT-1 0378-1119/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. PII: S 0378-111 9(01)00369-9 62 K.W. Sloop et al. / Gene 265 (2001) 61±69 patients, but individuals with this disease also typically display low or absent levels of LH and FSH (e.g. Duquesnoy et al., 1998; Wu et al., 1998; Pernasetti et al., 2000) and may exhibit late-onset ACTH de®ciency (e.g. Pernasetti et al., 2000). PROP-1 mutations include missense mutations that affect DNA interactions and frameshift mutations that result in truncated, inactive proteins. Panhypopituitarism and septo-optic dysplasia have been shown to result from an autosomal recessive mutation in the HESX1 gene (Dattani et al., 1998). This mutation causes an amino acid substitution that inhibits the DNA binding ability of this factor (Dattani et al., 1998). Humans with pituitary disease of unknown etiologic origin, such as patients with hypopituitarism and ectopically located posterior pituitary glands, have been hypothesized to have defects in pituitary regulatory genes such as PIT-1, PROP-1, and LHX3 (Hamilton et al., 1998; Pinto et al., 1999; Sloop et al., 2000a; Schmitt et al., 2000), but the molecular basis for these diseases remains unknown. Lhx3 (also known as P-Lim/LIM-3) is a LIM-HD transcription factor (Seidah et al., 1994; Bach et al., 1995; Zhadanov et al., 1995). This factor contains a DNA-binding homeodomain and two LIM domains that mediate protein/ protein interaction and trans-activation functions. Lhx3 is expressed in the embryonic rodent brain and spinal cord and later is restricted to the developing and adult pituitary gland (Seidah et al., 1994; Bach et al., 1995; Zhadanov et al., 1995). Studies in mice have demonstrated that it is essential for early pituitary structural development, for the speci®cation of motor neuron subtypes, and later for the differentiation of anterior and intermediate pituitary cell types (Sheng et al., 1996, 1997; Sharma et al., 1998). Lhx3 can activate transcription from anterior pituitary hormone gene promoters, either alone or in synergy with other pituitary regulatory proteins (e.g. Bach et al., 1995; Meier et al., 1999; Sloop et al., 1999). In humans, two LHX3 isoforms, LHX3a and LHX3b, are generated from the LHX3 gene (Sloop et al., 1999, 2000b; Schmitt et al., 2000). These isoforms possess identical LIM domains, homeodomains, and carboxyl termini but possess different amino termini that confer distinct functions upon the molecules (Sloop et al., 1999). Recently, mutations in the LHX3 gene have been shown to be associated with a novel, severe form of CPHD (Netchine et al., 2000). Two mutations in LHX3, a missense mutation of a tyrosine to a cysteine in the second LIM domain (LHX3aY111C, LHX3bY116C) and an intragenic deletion that results in a truncated protein lacking the DNA-binding homeodomain, were identi®ed. Patients with these mutations display retarded growth, pituitary hormone de®ciency, and abnormal neck and cervical spine development. The underlying mechanism of pituitary disease caused by mutations in the LHX3 gene has not been explored. In this study, we investigated the molecular basis of the LHX3 defects that cause human pituitary disease by comparing the gene regulatory properties of wild type and mutant LHX3 protein isoforms. 2. Materials and methods 2.1. Plasmid construction/mutagenesis The myc epitope-tagged LHX3a and LHX3b expression plasmids in the vector pcDNA3.1 have been described (Sloop et al., 1999). Y111/116C and Y111/116F mutations were introduced as described (Parker et al., 2000) using the following primers: 5 0 -cgcgcccaggacttcgtgtgccacctgcactgctttgcc-3 0 and 5 0 -ggcaaagcagtgcaggtggcacacgaagtcctgggcgcg-3 0 (Y111/116C); 5 0 -cgcgcccaggacttcgtgttccacctgcactgctttgcc-3 0 and 5 0 -ggcaaagcagtgcaggtggaacacgaagtcctgggcgcg-3 0 (Y111/116F). The LHX3 a/b DHD cDNA was generated using Expand High Fidelity DNA polymerase (Roche Biochemical) and the following primers: 5 0 -cgggatccatgctgctggaaacggggct-3 0 (LHX3a), 5 0 -cgggatccatggaggcgcgcggggagct-3 0 (LHX3b), and 5 0 -gcgaagcttggaccaggaaaggtgggagctgcttggcggtttcgtagtccgc-3 0 (LHX3 a/b DHD). Bacterial recombinant protein expression vectors for GSTLHX3 fusion proteins were constructed by cloning LHX3 cDNAs into pGEX-KT as described (Sloop et al., 1999). The integrity of all plasmids was con®rmed by DNA sequencing (Biochemistry and Molecular Biology, Indiana University School of Medicine). 2.2. In vitro transcription/translation Radiolabeled LHX3 proteins were synthesized in vitro from pcDNA3.1 expression vector substrates containing either wild type or mutant LHX3 cDNAs as described (Sloop et al., 1999). 2.3. Cell culture, transfection, and luciferase assays Mouse pituitary GHFT1-5 cells (Lew et al., 1993; kind gift of Dr. Pamela Mellon, University of California San Diego) and human embryonic kidney 293T cells were cultured as described (Sloop et al., 1999). Using the CalPhos system (Clontech), 1:5 £ 105 293T cells or 2:5 £ 105 GHFT1 cells/60 mm dish were transfected with calcium phosphate/DNA precipitates. Reporter plasmid 0.5 mg and 0.1±1.0 mg of expression vector were added per 60 mm dish, and all groups received equal ®nal DNA concentrations. Control cultures received empty expression vector DNA. The murine aGSU promoter luciferase plasmid (Roberson et al., 1994) was a kind gift of Dr. Richard Maurer (Oregon Health Sciences University). The Lhx3 consensus binding site reporter gene was previously described (Sloop et al., 1999). The rat Pit-1 and rat PRL promoter plasmids also have been described (Meier et al., 1999). Luciferase activity was measured 48 h after transfection as described (Meier et al., 1999). All assay points were performed in triplicate. Total cell protein was determined by the Bradford method (BioRad), and luciferase activity was normalized to protein concentration. K.W. Sloop et al. / Gene 265 (2001) 61±69 2.4. Statistical analysis Data points were compared using a one-tailed Student's ttest for paired samples using Sigma Plot 5.0 (Jandel Corp.). Values were considered signi®cantly different when P , 0:01. 2.5. Western analysis Western analyses of cells transfected with LHX3 expression vectors were performed as described (Meier et al., 1999). Mouse anti-myc 9E10 ascites ¯uid (Developmental Studies Hybridoma Bank, University of Iowa) was used at 1:5000. The secondary antibody was a goat anti-mouse/ horseradish peroxidase (Sigma) at 1:15000. Results were visualized using Lumi-Light PLUS chemiluminescence reagents (Roche) and Biomax MR ®lm (Kodak). 2.6. Recombinant protein preparation/electrophoretic mobility shift analysis Recombinant GST-LHX3 proteins were expressed in E. coli BL21 (DE3) pLysS and af®nity-puri®ed as previously described (Meier et al., 1999). Proteins were analyzed on 12% SDS-PAGE gels followed by staining with Coomassie brilliant blue. EMSAs were performed as described (Meier et al., 1999). Oligonucleotides representing the Lhx3 consensus binding site and murine aGSU promoter 2350 to 2323 bp element have been described (Sloop et al., 1999). 2.7. Protein/protein interaction assays Expression vectors containing mouse NLI and rat Pit-1 cDNA have been described (Meier et al., 1999). Labeled NLI and Pit-1 proteins were synthesized in vitro using TNT rabbit reticulocyte lysate reagents (Promega) and 35 S-methionine. Protein/protein interaction assays using labeled NLI or Pit-1 incubated with wild type and mutant GST-LHX3 fusion proteins were performed as previously described (Bach et al., 1995; Meier et al., 1999). 3. Results 3.1. Transcriptional properties of mutant LHX3 proteins Expression vectors encoding wild type LHX3a and LHX3b and LHX3aY111C, LHX3bY116C, LHX3aDHD, and LHX3bDHD mutant protein isoforms were constructed (Fig. 1A). In addition, because the amino acid residue at position 111/116 often is a phenylalanine in Lhx-class LIM-HD factors, expression vectors encoding LHX3aY111F and LHX3bY116F were generated to assess the general importance of this position (Fig. 1A). In vitro transcription/translation analysis demonstrated that the expression vectors produced proteins of the predicted relative molecular masses (Fig. 1B). We previously have demon- 63 strated that LHX3a isoform activates the promoter of the aGSU gene (Sloop et al., 1999). This gene encodes the common subunit of the LH, FSH, and TSH anterior pituitary hormones. By contrast, the LHX3b isoform has little or no effect on the induction of transcription from this promoter (Sloop et al., 1999). LHX3 expression vectors were cotransfected with an aGSU luciferase gene into either heterologous human embryonic kidney cells (293T) or mouse pituitary cells (GHFT1) and gene activity was recorded (Fig. 2A,B). In both cell types, LHX3a strongly induced the aGSU promoter, while LHX3b and all mutant derivatives of both LHX3 isoforms did not (P # 4 £ 1025 ). In 293T cells, the data points for LHX3bY116C and LHX3bY116F were slightly more than that for LHX3b (Fig. 2A), but this activity was not signi®cantly different from controls (P . 0:01), and this observation was not made in experiments using the GHFT1 pituitary cells (Fig. 2B). These data demonstrate that the LHX3 mutations associated with pituitary disease impair the ability of LHX3a to activate the aGSU promoter and do not inappropriately confer aGSU gene activation function to LHX3b. Experiments also were performed to test the ability of the mutant LHX3 proteins to activate complex pituitary promoters acting in combination with the Pit-1 transcription factor. Transient transfection experiments were performed using the PRL promoter in 293T and GHFT1 cells as described above. In these experiments, the LHX3aY111C and LHX3aY111F proteins were able to synergistically activate the PRL promoter with Pit-1 in 293T cells (Fig. 3A). In comparison to wild type LHX3a, the LHX3aY111C mutant exhibited signi®cantly reduced capacity to activate this gene in 293T cells (P 8 £ 1024 ) but LHX3aY111F did not (P 0:05). Similar observations were made in experiments using GHFT1 cells (Fig. 3B). The LHX3aDHD and LHX3bDHD mutants did not demonstrate synergy in these experiments (Fig. 3A,B). LHX3b did synergize with Pit-1 in these experiments, but the induction of the gene was low in comparison to assays using LHX3a (Fig. 3A,B). As observed with the aGSU promoter in 293T cells, luciferase activities in cotransfection experiments with LHX3bY116C or LHX3bY116F and Pit-1 were slightly higher than observed for LHX3b and Pit-1 (Fig. 3A) but, again, this observation was not made in experiments using GHFT1 cells (Fig. 3B). These results demonstrate that the Y116C mutation does not confer an increased capacity upon LHX3b to synergize with Pit-1. The ability of the mutant LHX3 proteins to activate a synthetic minimal reporter gene containing three copies of a Lhx3 binding site also was examined. Both LHX3a and LHX3b, with LHX3a having the greater activity, activate this reporter gene (Sloop et al., 1999). The Y111C and Y111F mutants of LHX3a both signi®cantly activated the synthetic reporter gene (Fig. 4A). The Y111C mutant displayed reduced activity compared to wild type (P 8 £ 1024 ), but LHX3aY111F did not (P 0:1). Similar effects were observed when the corresponding LHX3b 64 K.W. Sloop et al. / Gene 265 (2001) 61±69 Fig. 1. Wild type and mutant LHX3 proteins. (A) Schematic depiction of the LHX3a and LHX3b isoforms and of mutant molecules used in this study. Hatched regions, isoform-speci®c amino terminal domains; X, location of mutation; solid box, additional amino acids resulting from a frameshift caused by an intragenic deletion. (B) Radiolabeled wild type and mutant LHX3 proteins were generated from cDNA expression vectors by in vitro transcription/translation, separated by SDS electrophoresis, and dried gels were visualized by ¯uorography. The migration positions of protein standards (in kilodaltons) are shown. derivatives were tested (Fig. 4A). The mutant LHX3 proteins lacking the homeodomain were inactive. Western analysis was performed as a control to examine the levels of mutant and wild type proteins produced in transfected cells (Fig. 4B). In these experiments, it often was observed that LHX3aY111F was expressed at slightly reduced levels compared to wild type LHX3a. Together, these data indicate that the Y111/116C mutant LHX3 proteins have a reduced capacity to activate pituitary target genes and that the truncated mutants lacking the homeodomain are inactive. 3.2. DNA binding properties Recombinant derivatives of the wild type and mutant LHX3 proteins were generated as GST fusion proteins by expression in E. coli followed by af®nity puri®cation. EMSA analysis was performed using equivalent amounts of puri®ed proteins and radiolabeled Lhx3 DNA binding sites. LHX3a and LHX3aY111C displayed similar interaction with a consensus high-af®nity Lhx3 binding site (Fig. 5). By comparison to LHX3a, LHX3b has a reduced af®nity K.W. Sloop et al. / Gene 265 (2001) 61±69 65 family of factors (Bach, 2000), Pit-1 (Bach et al., 1995), MRG1 (Glenn and Maurer, 1999), and SLB (Howard and Maurer, 2000). To test the hypothesis that the Y111/116C mutation disrupts the LIM2 domain structure of LHX3 proteins and their binding to the partner proteins, we performed in vitro binding assays using puri®ed LHX3 proteins and NLI or Pit-1. Af®nity resins containing equivalent amounts of wild type and Y111/116C LHX3 isoforms were synthesized (Fig. 6E) and used as substrates in binding assays with radiolabeled NLI or Pit-1 ligands. Both the wild type and Y111/116C LHX3 proteins demonstrated signi®cant binding to NLI or Pit-1 (Fig. 6A/C,B/D, respectively). LHX3a, LHX3b, and the Y111/116C mutants displayed similar binding to NLI (Fig. 6A,C). However, the interac- Fig. 2. Mutations in LHX3 inhibit activation of the aGSU gene promoter. Human 293T cells (A) and mouse pituitary GHFT1 cells (B) were transiently transfected with an aGSU luciferase reporter gene and the indicated expression vectors. Promoter activity was assayed by measurement of luciferase activity after 48 h. Activities are mean [light units/10 s/mg total protein] of triplicate assays ^SEM. A representative experiment of at least ®ve experiments is depicted. for this class of DNA element (Sloop et al., 1999). As observed for LHX3a, the Y116C mutation did not affect the DNA binding capacity of LHX3b (Fig. 5). As expected, LHX3aDHD and LHX3bDHD proteins did not bind the tested DNA probes (Fig. 5). Similar data were obtained using the 2323 bp aGSU promoter Lhx3 binding element (data not shown). 3.3. Interaction of mutant LHX3 proteins with regulatory protein partners The two LIM domains of LIM-HD proteins such as Lhx3 allow interactions with partner proteins (reviewed in Bach, 2000). These proteins include the NLI/Ldb1/CLIM/Chip Fig. 3. Mutations in LHX3 impair synergistic induction of the PRL promoter. 293T (A) and GHFT1 (B) cells were transiently transfected with a PRL promoter reporter gene plasmid and wild type or mutant LHX3 and/or Pit-1 expression vectors. Luciferase activity was assayed after 48 h. Values are mean [light units/10 s/mg total protein] of triplicate assays ^SEM. A representative experiment of at least three experiments is shown. 66 K.W. Sloop et al. / Gene 265 (2001) 61±69 to date only a few mutations in this class of regulatory genes have been correlated with human diseases. For example, mutations in the LMX1B gene cause Nail-Patella syndrome (Dreyer et al., 1998) and Zhao et al. (1999) have suggested LHX8 as a candidate gene for a form of cleft palate in humans. Loss-of-function mutations of mouse LIM-HD genes often are lethal (e.g. reviewed in Hobert and Westphal, 2000), suggesting that severe mutations in the orthologous human genes might have similar effects. However, in contrast to the observation that mice lacking the Lhx3 gene die as neonates (Sheng et al., 1996), patients with LHX3 gene mutations survive and display a new form of CPHD (Netchine et al., 2000). In this study, we have demonstrated that the mutant LHX3 proteins have a reduced gene activation capacity. The Y111/116C mutant and the truncated LHX3 proteins are incapable of inducing the aGSU gene, which encodes an essential component of the three anterior pituitary glycopro- Fig. 4. Mutations in LHX3 reduce activation of a synthetic luciferase reporter gene containing Lhx3 binding sites. (A) Human 293T cells were transiently transfected with a Lhx3 reporter gene and the indicated expression vectors. Promoter activity was assayed by measurement of luciferase activity after 48 h. Activities are mean [light units/10 s/mg total protein] of triplicate assays ^SEM. A representative experiment of at least three experiments is depicted. (B) Western analysis using an anti-myc monoclonal antibody of cells transfected with myc epitope-tagged LHX3 expression vectors con®rmed expression of proteins. The migration positions of protein standards (in kilodaltons) are shown. tion of the Y111/116C proteins to Pit-1 was reduced compared to wild type LHX3 (Fig. 6B,D). These data indicate that the mutant LHX3 molecules retain their ability to interact with NLI and Pit-1, but the binding to Pit1 is reduced. 4. Discussion Although critical roles for LIM-HD genes in guiding the development of both vertebrate and invertebrate species have been demonstrated (reviewed in Dawid et al., 1998), Fig. 5. DNA binding of wild type and mutant LHX3 proteins. An EMSA using a Lhx3 binding site probe was performed. Radiolabeled probe was incubated with the indicated proteins and the resulting complexes were separated from free probe (F) by electrophoresis. LHX3 protein/DNA complexes are indicated by an arrow. Lane 1, probe alone; lane 2, GST as a negative control; lanes 3±8, recombinant LHX3 proteins expressed in E. coli. K.W. Sloop et al. / Gene 265 (2001) 61±69 67 Fig. 6. Interaction of wild type and mutant LHX3 proteins with protein partners. In vitro binding assays demonstrate interaction of wild type and mutant LHX3 proteins with NLI (A) and Pit-1 (B) proteins. Radiolabeled proteins were generated by translation in the presence of [ 35S]-methionine and incubated with the indicated GST fusion proteins or with GST alone as a control. After washing, equivalent samples of bound proteins were separated by electrophoresis and visualized by ¯uorography. The migration positions of molecular weight standards (in kilodaltons) are shown. The binding of NLI (C) and Pit-1 (D) proteins also was quanti®ed by scintillation counting. Bars indicate binding relative to that of LHX3a and represent the means of three independent experiments ^SEM. (E) Coomassie brilliant blue stain of a SDS-polyacrylamide gel analysis of GST-LHX3 fusion proteins used in interaction studies. The migration of molecular weight standards (in kilodaltons) is indicated. tein hormones and is an early marker for pituitary development. The truncated isoforms also are unable to activate the PRL gene. By contrast, the LHX3aY111C protein retains some ability to synergize with Pit-1 in activating PRL, although at a reduced level compared to wild type LHX3a. However, the fact that the Y111/116C mutation is associated with a similar disease to the more severe truncation mutation (Netchine et al., 2000) indicates that the Y111/116 residue is critical to overall LHX3 function. Indeed, a reduced capacity to activate certain target genes, rather than complete loss of activity, may be suf®cient to dramatically perturb pituitary development, especially target genes that are involved in critical regulatory steps of early pituitary ontogeny: a role that is predicted for LHX3 (Sheng et al., 1996, 1997). Modest changes in gene expression levels can have drastic consequences, especially 68 K.W. Sloop et al. / Gene 265 (2001) 61±69 if the affected gene is critical for the function of a cell or tissue. For example, a modest change in collagen type I gene expression is a key feature of ®brotic diseases (Jimenez et al., 1986). Maurer and colleagues have mapped a trans-activation function within the LIM domains of mouse Lhx3 (Glenn and Maurer, 1999). The reduced activity of the Y111/ 116C proteins, therefore, is consistent with a reduction in the ability of this domain to positively contribute to gene activation. It also is possible that Y111/116 is post-translationally modi®ed by phosphorylation and that such modi®cation is required for activity of the protein. The LIM domain is characterized by the presence of coordinated zinc ions at the base of each of the two `®ngers' (PerezAlvarado et al., 1994). In the Y111/116C mutant, the introduced cysteine may affect the structure of the LHX3 LIM2 motif by providing an inappropriate coordination position for the zinc ions that are integral to the LIM domain structure. Alternately, the introduced cysteine might form novel disul®de bridges that disrupt LHX3 structure. It is clear that the LIM domains are essential for Lhx-class protein functions. For example, Cheah and colleagues (2000) recently have demonstrated that mice homozygous for a Lhx1 allele encoding a protein with point mutations of the LIM domains are a phenocopy for Lhx1 null mice that display dramatic defects in head development. The LIM domains of LIM-HD proteins mediate protein/ protein interactions (reviewed in Bach, 2000). In our experiments, we demonstrated that LHX3 interactions with the NLI partner protein are maintained in the presence of the diseasecausing Y111/116C mutation. However, because LHX3 contains two LIM domains, both of which possess some independent ability to interact with known partners such as the NLI/CLIM proteins (e.g. Bach et al., 1997), disruption of LIM2 by the Y111/116C mutation might not affect overall recognition of such proteins by LHX3. It is possible, therefore, that the Y111/116C mutation prevents LIM2-dependent LHX3 interaction with other factors that are required for its activities in vivo. Binding of the Y111/116C mutant forms of LHX3 to Pit-1 is reduced. This observation may, in part, explain the reduction in ability of LHX3aY111C to activate the PRL promoter in the presence of Pit-1. We conclude that the mutant forms of the LHX3a and LHX3b neuroendocrine transcription factor proteins display impaired abilities to activate pituitary hormone gene promoters. Whereas the Y111/116C mutant LHX3 proteins exhibit similar DNA binding properties to the wild type proteins on known LHX3 DNA-binding elements, the truncated LHX3a and LHX3b proteins lacking the homeodomain do not bind DNA. The Y111/116C mutant LHX3 proteins retain some ability to interact with the NLI and Pit-1 partner proteins. To fully understand the mechanism by which these mutations cause complex disease, a more complete identi®cation of the target genes and partner proteins for the LHX3 proteins during pituitary and motor neuron development will be required. Acknowledgements We are grateful to Drs R. Maurer and P. Mellon for materials and J. Bridwell and A. Hartman for advice. Supported by grants to SJR from the National Science Foundation and the NRICGP/USDA. References Bach, I., 2000. The LIM domain: regulation by association. Mech. Dev. 91, 5±17. Bach, I., Rhodes, S.J., Pearse II, R.V., Heinzel, T., Gloss, B., Scully, K.M., Sawchenko, P.E., Rosenfeld, M.G., 1995. P-Lim, a LIM homeodomain factor, is expressed during pituitary organ and cell commitment and synergizes with Pit-1. Proc. Natl. Acad. Sci. USA 92, 2720±2724. Bach, I., CarrieÁre, C., Ostendorff, H.P., Anderson, B., Rosenfeld, M.G., 1997. A family of LIM domain-associated cofactors confer transcriptional synergism between LIM and Otx homeodomain proteins. Genes Dev. 11, 1370±1380. Burrows, H.L., Douglas, K.R., Seasholtz, A.F., Camper, S.A., 2000. Genealogy of the anterior pituitary gland: tracing a family tree. Trends Endocrinol. Metab. 10, 343±352. Cheah, S.S., Kwan, K.M., Behringer, R.R., 2000. Requirement of LIM domains for LIM1 function in mouse head development. Genesis 27, 12±21. Cohen, L.E., Wondisford, F.E., Salvatoni, A., Maghnie, M., Brucker-Davis, F., Weintraub, B.D., Radovick, S., 1995. A `hot spot' in the Pit-1 gene responsible for combined pituitary hormone de®ciency: clinical and molecular correlates. J. Clin. Endocrinol. Metab. 80, 679±684. Dattani, M.T., Martinez-Barbera, J.P., Thomas, P.Q., Brickman, J.M., Gupta, R., Martensson, I.L., Toresson, H., Fox, M., Wales, J.K., Hindmarsh, P.C., Krauss, S., Beddington, R.S., Robinson, I.C., 1998. Mutations in the homeobox gene HESX1/Hesx1 associated with septo-optic dysplasia in human and mouse. Nat. Genet. 19, 125±133. Dawid, I.B., Breen, J.J., Toyama, R., 1998. LIM domains: multiple roles as adapters and functional modi®ers in protein interactions. Trends Genet. 14, 295±306. Dreyer, S., Zhou, G., Baldini, A., Winterpacht, A., Zabel, B., Cole, W., Johnson, R., Lee, B., 1998. Mutations in LMX1B cause abnormal skeletal patterning and renal dysplasia in nail patella syndrome. Nat. Genet. 19, 47±50. Duquesnoy, P., Roy, A., Dastot, F., Ghali, I., Teinturier, C., Netchine, I., Cacheux, V., Hafez, M., Salah, N., Chaussain, J.L., Goossens, M., Bougneres, P., Amselem, S., 1998. Human Prop-1: cloning, mapping, genomic structure. Mutations in familial combined pituitary hormone de®ciency. FEBS Lett. 437, 216±220. Glenn, D.J., Maurer, R.A., 1999. MRG1 binds to the LIM domain of Lhx2 and may function as a coactivator to stimulate glycoprotein hormone alpha-subunit gene expression. J. Biol. Chem. 274, 36159±36167. Hamilton, J., Chitayat, D., Blaser, S., Cohen, L.E., Phillips, J.A., 3rd, ., Daneman, D., 1998. Familial growth hormone de®ciency associated with MRI abnormalities. Am. J. Med. Genet. 80, 128±132. Hobert, O., Westphal, H., 2000. Functions of LIM-homeobox genes. Trends Genet. 16, 75±83. Howard, P.W., Maurer, R.A., 2000. Identi®cation of a conserved protein that interacts with speci®c LIM homeodomain transcription factors. J. Biol. Chem. 275, 13336±13342. Irie, Y., Tatsumi, K., Ogawa, M., Kamijo, T., Preeyasombat, C., Suprasongsin, C., Amino, N., 1995. A novel E250X mutation of the Pit-1 gene in a patient with combined pituitary hormone de®ciency. Endocr. J. 42, 351±354. Jacobson, E.M., Li, P., Leon-del-Rio, A., Rosenfeld, M.G., Aggarwal, A.K., 1997. Structure of Pit-1 POU domain bound to DNA as a K.W. Sloop et al. / Gene 265 (2001) 61±69 dimer: unexpected arrangement and ¯exibility. Genes Dev. 11, 198± 212. Jimenez, S.A., Feldman, G., Bashey, R.I., Bienkowski, R., Rosenbloom, J., 1986. Co-ordinate increase in the expression of type I and type III collagen genes in progressive systemic sclerosis ®broblasts. Biochem. J. 237, 837±843. Lew, D., Brady, H., Klausing, K., Yaginuma, K., Theill, L., Stauber, C., Karin, M., Mellon, P.L., 1993. GHF-1-promoter-targeted immortalization of a somatotropic progenitor cell results in dwar®sm in transgenic mice. Genes Dev. 7, 683±693. Meier, B.C., Price, J.R., Parker, G.E., Bridwell, J.L., Rhodes, S.J., 1999. Characterization of the porcine Lhx3/LIM-3/P-Lim LIM homeodomain transcription factor. Mol. Cell. Endocrinol. 147, 65±74. Netchine, I., Sobrier, M.L., Krude, H., Schnabel, D., Maghnie, M., Marcos, E., Duriez, B., Cacheux, V., Moers, A., Goossens, M., Gruters, A., Amselem, S., 2000. Mutations in LHX3 result in a new syndrome revealed by combined pituitary hormone de®ciency. Nat. Genet. 25, 182±186. Parker, G.E., Sandoval, R.M., Feister, H.A., Bidwell, J.P., Rhodes, S.J., 2000. The homeodomain coordinates nuclear entry of the Lhx3 neuroendocrine transcription factor and association with the nuclear matrix. J. Biol. Chem. 275, 23891±23898. Parks, J.S., Brown, M.R., 1999. Transcription factors regulating pituitary development. Growth Horm. IGF Res. 9, 2±8. Perez-Alvarado, G.C., Miles, C., Michelsen, J.W., Louis, H.A., Winge, D.R., Beckerle, M.C., Summers, M.F., 1994. Structure of the carboxy-terminal LIM domain from the cysteine rich protein CRP. Nat. Struct. Biol. 1, 388±398. PfaÈf¯e, R., DiMattia, G., Parks, J., Brown, M., Wit, J., Jansen, M., Van der Nat, H., Van den Brande, J., Rosenfeld, M.G., Ingraham, H.A., 1992. Mutation of the POU-speci®c domain of Pit1 and hypopituitarism without pituitary hypoplasia. Science 257, 1118±1121. Pernasetti, F., Milner, R., Al Ashwal, A., Zegher, F., Chavez, V., Muller, M., Martial, J., 1998. Pro239Ser: a novel recessive mutation of the Pit-1 gene in seven middle eastern children with growth hormone, prolactin, and thyrotropin de®ciency. J. Clin. Endocrinol. Metab. 83, 2079±2083. Pernasetti, F., Toledo, S., Vasilyev, V., Hayashida, C., Cogan, J., Ferrari, C., Lourenco, D., Mellon, P., 2000. Impaired adrenocorticotropin-adrenal axis in combined pituitary hormone de®ciency caused by a two-base pair deletion (301±302delAG) in the prophet of Pit-1 gene. J. Clin. Endocrinol. Metab. 85, 390±397. Pinto, G., Netchine, I., Sobrier, M.L., Brunelle, F., Souberbielle, J.C., Brauner, R.J., 1999. Pituitary stalk interruption syndrome: a clinical-biological-genetic assessment of its pathogenesis. J. Clin. Endocrinol. Metab. 82, 3450±3454. Roberson, M.S., Schoderbek, W.E., Tremml, G., Maurer, R.A., 1994. Acti- 69 vation of the glycoprotein hormone alpha subunit promoter by a LIMhomeodomain transcription factor. Mol. Cell. Biol. 14, 2985±2993. Schmitt, S., Biason-Lauber, A., Betts, D., Schoenle, E.J., 2000. Genomic structure, chromosomal localization, and expression pattern of the human LIM-homeobox3 (LHX 3) gene. Biochem. Biophys. Res. Commun. 274, 49±56. Seidah, N.G., Barale, J.C., Marcinkiewicz, M., Mattei, M.G., Day, R., ChreÂtien, M., 1994. The mouse homeoprotein mLIM-3 is expressed early in cells derived from the neuroepithelium and persists in adult pituitary. DNA Cell Biol. 13, 1163±1180. Sharma, K., Sheng, H.Z., Lettieri, K., Li, H., Karavanov, A., Potter, S., Westphal, H., Pfaff, S.L., 1998. LIM homeodomain factors Lhx3 and Lhx4 assign subtype identities for motor neurons. Cell 95, 817±828. Sheng, H.Z., Zhadanov, A.B., Mosinger Jr, B., Fujii, T., Bertuzzi, S., Grinberg, A., Lee, E.J., Huang, E.J., Huang, S.P., Mahon, K.A., Westphal, H., 1996. Speci®cation of pituitary cell lineages by the LIM homeobox gene Lhx3. Science 272, 1004±1007. Sheng, H.Z., Moriyama, K., Yamashita, T., Li, H., Potter, S.S., Mahon, K.A., Westphal, H., 1997. Multistep control of pituitary organogenesis. Science 278, 1809±1812. Sloop, K.W., Meier, B.C., Bridwell, J.L., Parker, G.E., Schiller, A.M., Rhodes, S.J., 1999. Differential activation of pituitary hormone genes by human Lhx3 isoforms with distinct DNA binding properties. Mol. Endocrinol. 13, 2212±2225. Sloop, K.W., Walvoord, E.C., Showalter, A.D., Pescovitz, O.H., Rhodes, S.J., 2000a. Molecular analysis of LHX3 and PROP-1 in pituitary hormone de®ciency patients with posterior pituitary ectopia. J. Clin. Endocrinol. Metab. 85, 2701±2708. Sloop, K.W., Showalter, A.D., Von Kap-Herr, C., Pettenati, M.J., Rhodes, S.J., 2000b. Analysis of the human LHX3 neuroendocrine transcription factor gene and mapping to the subtelomeric region of chromosome 9. Gene 245, 237±243. Tatsumi, K., Miyai, K., Notomi, T., Kaibe, K., Amino, N., Mizuno, Y., Kohno, H., 1992. Cretinism with combined hormone de®ciency caused by a mutation in the Pit-1 gene. Nat. Genet. 1, 56±58. Wu, W., Cogan, J., PfaÈf¯e, R., Dasen, J., Frisch, H., O'Connell, S., Flynn, S., Brown, M., Mullis, P., Parks, J., Phillips 3rd, J., Rosefeld, M.G., 1998. Mutations in PROP1 cause familial combined pituitary hormone de®ciency. Nat. Genet. 8, 147±149. Zhadanov, A.B., Bertuzzi, S., Taira, M., Dawid, I.B., Westphal, H., 1995. Expression pattern of the murine LIM class homeobox gene Lhx3 in subsets of neural and neuroendocrine tissues. Dev. Dyn. 202, 354±364. Zhao, Y., Guo, Y.J., Tomac, A.C., Taylor, N.R., Grinberg, A., Lee, E.J., Huang, S., Westphal, H., 1999. Isolated cleft palate in mice with a targeted mutation of the LIM homeobox gene Lhx8. Proc. Natl. Acad. Sci. USA 96, 15002±15006.